Re: mege databases of multiple countries over several years [message #20487 is a reply to message #20486] |
Wed, 11 November 2020 10:53 |
Bridgette-DHS
Messages: 3214 Registered: February 2013
|
Senior Member |
|
|
Following is another response from DHS Research & Data Analysis Director, Tom Pullum:
I suggest that you first select just one of these Cameroon surveys and try to identify a relationship between the predictors of interest (tobacco use, etc.) and the outcomes (stunting, underweight, wasted), including controls for other known predictors of the outcomes. You can do most of this with just the KR file, for children who are living with the mother (b9=0). Linking the child with the father in the MR file is not trivial. Stata code for stunting, etc. is available on the website (see the GitHub link). Good luck....
|
|
|